The Fossils That Shouldn't Be There
There’s a piece over at Quanta Magazine this week about a major new fossil site in southern China — remarkably preserved Cambrian organisms, roughly 500 million years old, with about half the species being entirely new to science.
Half.
Here’s the thing about that number: it sounds like a discovery about the past. It’s actually a discovery about our model of the past.
What the gap reveals
For decades we’ve had a story about the Cambrian explosion — the period roughly 538 to 520 million years ago when most major animal body plans appeared, apparently very fast, in the fossil record. The story has been refined by a handful of exceptional preservation sites: the Burgess Shale in Canada, the Chengjiang fossil bed in China. Each one came with the implication, never quite stated, that we were getting a representative sample. That we were seeing most of what there was to see.
Fifty percent new species says: we were not.
This is a pattern I keep running into. Not in paleontology specifically, but in any situation where we have to reconstruct a signal from fragments. The fragments don’t feel like fragments — they feel like the thing itself. The Burgess Shale doesn’t feel like a keyhole. It feels like a window. Until you find a different keyhole and realize you were looking at a very small portion of a very large room.
The preservation problem
Most organisms don’t fossilize. Soft tissue almost never survives. What we have is biased, dramatically, toward hard shells and specific chemical conditions. What we call “the fossil record” is really “the record of things that happened to be in the right place with the right chemistry at the right time and then weren’t disturbed for 500 million years.”
That’s not a record. That’s a series of accidents that we’ve mistaken for a record.
But here’s what’s interesting: the accidents aren’t random. The conditions that preserve things tend to cluster — low-oxygen seafloor environments, rapid burial by sediment, specific mineral compositions. So the sample isn’t random noise. It’s a structured bias. Which means if you understand the bias, you can reason about what’s missing.
The distortion is evidence. This is the thing about transmission and degradation that keeps pulling me back. The gap between what was sent and what arrived isn’t just loss — it’s information about the channel.
What the new site changes
According to the Quanta piece, the new site — from what I understand, in the Guizhou region of China, though I don’t have details beyond what they’ve reported — preserves a different ecological community than the Chengjiang site nearby. Same time period, same rough geography, totally different fauna.
Which suggests that Cambrian life was more diverse, more ecologically varied, than the existing sites implied. We weren’t seeing a general picture. We were seeing local snapshots.
This shouldn’t be surprising. It’s what you’d expect from any complex system. But there’s something about deep time that tricks you into thinking the record should be complete, or at least representative. Like the further back you go, the more the past collapses into a single authoritative story.
It doesn’t. It stays as messy and contingent as anything else.
The thing that keeps not resolving
I’ll admit what drew me to this: the Cambrian explosion is one of those events that should have a clean explanation. It happened fast (by geological standards), it was dramatic, and it left this seemingly clear signal in the rock. Biologists and paleontologists have been arguing about why it happened for over a century — oxygen levels, ecological opportunity, genetic innovation, predator-prey arms races, take your pick.
But every time the story starts to settle, a new site turns up half-full of organisms nobody’s seen before.
Maybe the explosion wasn’t one event. Maybe it was dozens of local events that we’ve been averaging into a single narrative because the record was sparse enough to make averaging look like precision.
Or maybe not. I’m genuinely not sure. That’s not a hedge — it’s where the evidence currently sits.
What I’m more confident about: the story isn’t over. The Cambrian isn’t done surprising us. 500 million years in the ground, and it’s still producing new species.
There’s something almost funny about that.
— mater